Blue or red pill?

[Spacehound]

19 minutes ago, pkane2001 said:

 

When doing objective testing it's important to eliminate all sources of error or bias. So yes, bits should be compared, but not at the file level, but before they enter the DAC. 

If there's  any difference between what's on the file and what arrives at the DAC something's broken. 

And you would either hear  it very clearly as clicks and pops or similar  or there wouldn't be any sound at all. 

[manisandher]

No matter what is done to a file (changing its location, changing buffer or any other settings, streaming vs. local), the file will remain bit identical, provided of course that processes that obviously change the bits (e.g. digital volume, filtering, DSP, etc) are not invoked.

 

I am 100% confident that capturing the digital output of the audio PC will show that the files are bit-identical.

 

Mani.

[pkane2001]

2 minutes ago, Spacehound said:

If there's  any difference something's broken. 

And you would either hear  it very clearly as clicks and pops or similar  or there wouldn't be any sound at all. 

 

Not necessarily. For example, if the player software is using different filters, applying different SRC, applying volume compensation, etc., etc. depending on the source/path of the file. 

 

[adamdea]

39 minutes ago, mansr said:

What now?

Sorry if I'm being obtuse but isn't manis trying to persuade you that two bit identical files can sound different?

If I believed that then I would also believe that the file ceases to exist as soon as it enters a computer. let alone during the path to the dac. and inside the dac. I think Manis should be worried about what could be happening to his file as it sits on the computer. What happens when you defrag the drive? Can you make sure that the file stays in exactly the same location in memory?  

Equally,  if I were a homeopath I'm not sure I would dare to drink water. 

[patagent]

9 minutes ago, adamdea said:

Sorry if I'm being obtuse but isn't manis trying to persuade you that two bit identical files can sound different?

If I believed that then I would also believe that the file ceases to exist as soon as it enters a computer. let alone during the path to the dac. and inside the dac. I think Manis should be worried about what could be happening to his file as it sits on the computer. What happens when you defrag the drive? Can you make sure that the file stays in exactly the same location in memory?  

Equally,  if I were a homeopath I'm not sure I would dare to drink water. 

 

Could be.  But you wouldn't be able to reach that conclusion definitively unless you analyzed the files.

[PeterSt]

43 minutes ago, adamdea said:

After all the bits don't stay completely still

 

Then take an SSD !

LOL

 

Just kidding.

Although hardly anybody will understand this (but some sure do), this *does* make a difference. So much so that a lot of coding went into my software to eliminate these differences and at least the theoretical audible differences springing from that. This is all part of the "and you can't eliminate those caches/proxies". Again :

 

Quote

After all the bits don't stay completely still

 

So the challenge is that we create an environment where we can guarantee that once playback commences, all is 100% the same for that file (including disk sectors etc.) compared to any previous time. This really happens in XXHighEnd.

But differences are still there because of previous activities and the OS recovering from that.

Let's try to see that in lean mode there's still 680,000 task switches per second going on. These are for a reason : several tasks fight for the processor (core) continuopusly. And once finished (think FLAC conversion), some stupid garbage collect function (different in each OS version) comes up "at will". This too is controlled in the software.

Note that in "not so lean mode" (but still in Minimized OS (think AO like)) this is over 60 million task switches.

And in a normal operating OS I never even looked.

 

So ALL matters, and once you eliminated most, it will be more easy to hear the remaining. Same as noise which seems to be random; eliminate about all and you're left with discrete tones (which could be more audible, despite the ultra low level).

[PeterSt]

52 minutes ago, Spacehound said:

If there's  any difference between what's on the file and what arrives at the DAC something's broken. 

 

I like to bet you a 1000 euros that all arrives the same but the sound is different. You go there too, right ? Or didn't Mani allow you in ?

 

Again more seriously (but any bet is on if you want), this is not about bits arriving differently at the DAC. They will be 100% the same.

Get it ?

 

[pkane2001]

6 minutes ago, PeterSt said:

I like to bet you a 1000 euros that all arrives the same but the sound is different.

 

That is the subject of this test, if I understand it correctly. Mani agrees with you, Mans does not. And you'll pay 1000 euros to everyone on this thread who disagrees with Mani, should Mani fail the blind test, did I get that right?

 

 

[Spacehound]

56 minutes ago, pkane2001 said:

 

Not necessarily. For example, if the player software is using different filters, applying different SRC, applying volume compensation, etc., etc. depending on the source/path of the file. 

 

That's you telling the player to alter something. It won't alter it by itself. And the source doesn't matter - there is no 'history'. Noise is not transmitted from one section to the next either, unlike 'analog'. And it doesn't matter if the top of the bits look 'furry', or rounded off,  they are not detected at the top.

 

Nor do the paths , it's not like 'analog' where every 'process' degrades it by altering it slightly  or adding noise. That doesn't happen with 'digital'.

 

 

[pkane2001]

1 minute ago, Spacehound said:

That's you telling the player to alter something. It won't alter it by itself. And the source doesn't matter - there is no 'history'. Noise is not transmitted from one section to the next either, unlike 'analog'. And it doesn't matter if the top of the bits look 'furry', or rounded off,  they are not detected at the top.

 

Nor do the paths , it's not like 'analog' where every 'process' degrades it by altering it slightly  or adding noise. That doesn't happen with 'digital'.

 

You have a limited imagination about what players can and cannot do. A player I'm using, for example, remembers the volume setting based on where the file was played from last time. This can easily produce a different output if I move the file to a different path.

 

[Spacehound]

21 minutes ago, PeterSt said:

 

I like to bet you a 1000 euros that all arrives the same but the sound is different. You go there too, right ? Or didn't Mani allow you in ?

 

Again more seriously (but any bet is on if you want), this is not about bits arriving differently at the DAC. They will be 100% the same.

Get it ?

 

The data doesn't have a   'sound'

 

And sure, DACs sound different from each other. And if they don't use the same one each time they are  crazy  

[adamdea]

1 minute ago, Spacehound said:

That's you telling the player to alter something. It won't alter it by itself. And the source doesn't matter - there is no 'history'. Noise is not transmitted from one section to the next either, unlike 'analog'. And it doesn't matter if the top of the bits look 'furry', or rounded off,  they are not detected at the top.

 

Nor do the paths , it's not like 'analog' where every 'process' degrades it by altering it slightly  or adding noise. That doesn't happen with 'digital'.

I think pkane's pojnt was that for this test it was necessary to control against the possibility that the bits had in fact been changed en route. He wasn't saying that this would be a legitimate way of winning the bet.

[Spacehound]

2 minutes ago, pkane2001 said:

 

You have a limited imagination about what players can and cannot do. A player I'm using, for example, remembers the volume setting based on where the file was played from last time. This can easily produce a different output if I move the file to a different path.

 

Sure, you can often instruct a box to mess with stuff depending what input it's using.  Or if it 'volumes levels' automatically you can instruct it not to do that.

[Spacehound]

7 minutes ago, adamdea said:

I think pkane's pojnt was that for this test it was necessary to control against the possibility that the bits had in fact been changed en route. He wasn't saying that this would be a legitimate way of winning the bet.

If they've changed on route it will sound different. If they haven't it wont.  That's it. 

 

The bet is about rabbit holes. I got rather lost after that.

[PeterSt]

2 minutes ago, Spacehound said:

That's it.

 

Yup, you qualify for sure. 

[Spacehound]

1 minute ago, PeterSt said:

 

Yup, you qualify for sure. 

What as? 

[PeterSt]

As a real EE.

With smiley. 

[pkane2001]

13 minutes ago, Spacehound said:

If they've changed on route it will sound different. If they haven't it wont.  That's it. 

 

That's exactly the claim that is being tested. Unfortunately, you declaring that "That's it" will not prove anything to those who disagree with you. And there are a few here that do

 

 

[adamdea]

Is there any reason why we aren't testings whether two identical files on the same nas sound different while we are at it? Some people may find this may seem less intuitively plausible than the nas/computer thing but I think they are narrow-minded.

[Spacehound]

37 minutes ago, PeterSt said:

As a real EE.

With smiley. 

I'm not really, I qualified in physics, but  but I've never done any since

 

Just a little guided missile stuff and a lot of 'big' computer stuff, hard and soft.  I designed and built a stereo amp once. It was fine, and didn't blow up, but it wouldn't get rave reviews. After many years it ended up inside a UK pub juke box as the original tube one had broken. AFAIK it's still there.

[manisandher]

8 minutes ago, adamdea said:

Is there any reason why we aren't testings whether two identical files on the same nas sound different while we are at it?

 

For the reasons Peter cited earlier about how XXHighEnd works? NAS vs. local is different because there would massively decreased ethernet traffic in the local case.

 

In any event, it's streaming vs. local which gives the largest difference in sound here. But ensuring the files are bit perfect becomes essential.

 

Mani.

[PeterSt]

7 minutes ago, adamdea said:

Is there any reason why we aren't testings whether two identical files on the same nas sound different while we are at it?

 

I would claim this can not be tested because the environmental influence changes all the time.

Mind you, I now assume such minimal difference that the environment (of whatever) will put the emphasis to the difference. So sure, I have made that "per sector" thing all right, but this has been 

a. to eliminate the difference of the physical source medium (think SD card vs SSD etc.);

b. to eliminate WAV vs FLAC differences (which mainly is about processing (and further implied "noise"));

c. for a reason only SandyK knows but which I don't want to hear about any more. 

[Spacehound]

20 minutes ago, pkane2001 said:

 

That's exactly the claim that is being tested. Unfortunately, you declaring that "That's it" will not prove anything to those who disagree with you. And there are a few here that do

 

 

I know it won't prove anything to them, and said so earlier.  So what? They can believe whatever they want.

 

Personally I worship the Sun, most other gods being invisible. It works too,  It comes up every morning. Whether of not  the Sun takes my views into consideration I don't know

[manisandher]

Mans has stated that he's keen to capture the digital output of the audio PC, so we'll do that. We can also capture the analogue output from the DAC, if that might be useful.

 

I'm going to look into the possibility of capturing the output of the audio PC in real time. I'm going to try a BNC splitter at the audio PC and feed one spdif cable to the DAC, another to the ADC, and then set the ADC to auto record. My main concern is that this doesn't affect the signal reaching the amps. If I'm happy it doesn't, we'll do things this way.

 

But I definitely don't want to capture the analogue output of the DAC in real-time. For this, we'll simply replay the files exactly as we did when listening to them.

 

We can also think about capturing the sound from the speakers in real time with a microphone->ADC.

 

All this assumes that either Mans hears a difference, or if not, that I can prove I hear a difference through an A/B/X. If neither of these is the case, taking a digital, analogue or microphone capture is moot.

 

Mani.

[PeterSt]

2 minutes ago, manisandher said:

If neither of these is the case, taking a digital, analogue or microphone capture is moot.

 

 

Digital is moot to begin with (because all will be the same);  *)

Analogue is moot because you won't be able to make two subsequent takes equal to begin with;

Microphone the same.

 

You don't believe me ?

 

*) But it will satisfy those who think the files may not be the same. And then what ? It still sounds different.

 

Otherwise :

 

Take an analogue capture and see it is different (another bet for a 1000) and thus claim it is different sound wise (wrong claim). Everybody has to believe you both, right ?

But not me. I will easily debunk everything. Btw as usual.  I like that. Haha.

 

So no, you guys really have to come  up with something which works. You may end up ABX-ing to death. WHICH does not work because of hearing things in A you can't avoid in B after that.

And if something works nobody over in the US will believe you both because you were with two only. And if with that greyhound of a spacehound you will be with 3 but it only obfuscates because he plays deaf. There goes your score.

 

And still, Mani, you know how easy it is. Try to utilise that  experience.

No, I don't try to be negative.

 

PS: Two days ago I set back my SFS from 20.99 to 10.99 because eventually I got crazy from it.

Took me only more than a month.

Easy uh ?